X pauses until mouse is moved

[Kevin Oberman]

> From: Joe Marcus Clarke <marcus at FreeBSD.org>
> Date: Wed, 26 Mar 2008 01:47:38 -0400
> 
> On Tue, 2008-03-25 at 17:38 -0400, Joe Marcus Clarke wrote:
> > On Tue, 2008-03-25 at 14:31 -0700, Kevin Oberman wrote:
> > > > From: Joe Marcus Clarke <marcus at FreeBSD.org>
> > > > Date: Tue, 25 Mar 2008 12:07:00 -0400
> > > > Sender: owner-freebsd-x11 at freebsd.org
> > > > 
> > > > This problem was originally reported on this list on March 5
> > > > (http://lists.freebsd.org/pipermail/freebsd-x11/2008-March/006077.html).
> > > > I am now seeing this on my RELENG_7 and -CURRENT boxes.  Basically, all
> > > > interaction with X is temporarily suspended until the mouse is moved.
> > > > This only occurs when using /dev/sysmouse (thus when moused is enabled).
> > > > If I disabled moused, and use /dev/psm0 directly, the problem goes away.
> > > > 
> > > > My i386 RELENG_7 machine was working fine until I updated to:
> > > > 
> > > > FreeBSD shumai.marcuscom.com 7.0-STABLE FreeBSD 7.0-STABLE #17: Mon Mar
> > > > 24 15:32:39 EDT 2008
> > > > marcus at shumai.marcuscom.com:/build/obj/build/src/sys/SHUMAI  i386
> > > > 
> > > > Prior to that I was running FreeBSD 7.0-STABLE #16: Sat Mar  8 20:07:36
> > > > EST 2008.
> > > > 
> > > > Also prior to that I had the xorg-server update that was supposed to fix
> > > > jerky mouse movement.  That didn't seem to trigger this problem.  I
> > > > thought it might have been related to the recent moused fix in RELENG_7,
> > > > so I backed out the moused.c changes, but the problem persists.  I also
> > > > backed out the recent X mouse driver VT switch fix, but the problem
> > > > persists.
> > > > 
> > > > At least two other users have described similar problems.  Any
> > > > suggestions on what may be causing this?  The only difference I spot in
> > > > dmesg relates to CPU clock speed (off by 1/100 of a MHz).  The working
> > > > version of FreeBSD had:
> > > > 
> > > > CPU: Intel(R) Xeon(R) CPU            5140  @ 2.33GHz (2327.52-MHz
> > > > 686-class CPU)
> > > > 
> > > > The current version has:
> > > > 
> > > > CPU: Intel(R) Xeon(R) CPU            5140  @ 2.33GHz (2327.51-MHz
> > > > 686-class CPU)
> > > > 
> > > > A full (current) dmesg can be found at
> > > > http://www.marcuscom.com/downloads/dmesg.shumai .
> > > 
> > > I am seeing about the same thing here. My system is running:
> > > FreeBSD slan.es.net 7.0-STABLE FreeBSD 7.0-STABLE #2: Mon Mar 17 21:39:01 PDT 2008     root at slan.es.net:/usr/obj/usr/src/sys/IBM-T43  i386
> > > 
> > > What is possibly notable is that I only started seeing this problem
> > > yesterday, right after upgrading to Gnome 2.22. It looks like the Gnome
> > > upgrade triggered something, possibly an interaction with the moused,
> > > sysmouse, or xf86-input-mouse.
> > > 
> > > The system is a T43 using the internal keyboard and TrackPoint(tm).
> > > 
> > > The Gnome upgrade was pretty smooth with everything building, but
> > > portupgrade complaining about some dependency loops. (I'll report about
> > > this to the Gnome list.)
> > 
> > If you follow /usr/ports/UPDATING these loops will go away.  There are
> > no real loops.  It may require you to run pkgdb -fF a few times, though.
> > 
> > > 
> > > This is more than a bit annoying. It also impacts menus and scroll
> > > bars. I plan to drop back to my backup from before the Gnome upgrade.
> > > 
> > > I can make config, xorg.conf, and dmesg available, but I can't see
> > > anything odd there.
> > 
> > I have a GNOME 2.22 machine with a USB mouse running -CURRENT from March
> > 6 which is not seeing this problem.  I've also seen the issue running a
> > failsafe xterm with no window manager, and no gnome-session.  If the
> > GNOME update introduced this problem, I'm not sure how.
> 
> The problem is ConsoleKit.  If console-kit-daemon is running, then
> moused an X get all screwy.  I think this only affects GDM users, and
> may be fixed with my recent GDM commit.  I am now running X with CK and
> moused with no pausing.  Can people seeing this problem confirm if
> you're using GDM?  If not, try starting X AFTER console-kit-daemon, and
> see if that helps.  Note: bland has most likely fixed this bug in very
> recent -CURRENT.

Afraid not. I don't use GDM and I did not have console-kit-daemon
running but still saw the problem. (That was my first guess as to the
cause.) 

I'm not sure what starts console-kit-daemon, but it is running just fine
right now. The only change I made to get things working was to remove
the hald_enable from rc.conf.
-- 
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman at es.net			Phone: +1 510 486-8634
Key fingerprint:059B 2DDF 031C 9BA3 14A4  EADA 927D EBB3 987B 3751
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 224 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-x11/attachments/20080326/9e9f4d98/attachment.pgp